[OpenAFS] Re: Strange problem with OpenAFS on Virtual Machine

Claudio Prono claudio.prono@atpss.net
Fri, 09 Jul 2010 18:20:25 +0200


This is a multi-part message in MIME format.
--------------050101090601000706020903
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

Derrick Brashear ha scritto:
> On Fri, Jul 9, 2010 at 12:03 PM, Andrew Deason <adeason@sinenomine.net> wrote:
>   
>> On Fri, 9 Jul 2010 11:51:01 -0400
>> Derrick Brashear <shadow@gmail.com> wrote:
>>
>>     
>>>> Could we not just try to kmalloc if vmalloc fails repeatedly?
>>>>         
>>> we don't elsewhere allocate sizes > PAGESIZE from kmalloc, and this
>>> definitely would be.
>>>       
>> Yes, but why not? kmalloc can alloc blocks bigger than PAGESIZE, I
>> thought (it goes up to 128k or something like that?).
>>     
>
>   
But... there is something strange...

afs-test:~ # cat /proc/meminfo
MemTotal:      1034864 kB
MemFree:        821272 kB
Buffers:         52100 kB
Cached:          79408 kB
SwapCached:          0 kB
Active:         102700 kB
Inactive:        84524 kB
HighTotal:      131008 kB
HighFree:          252 kB
LowTotal:       903856 kB
LowFree:        821020 kB
SwapTotal:     1052216 kB
SwapFree:      1052216 kB
Dirty:            1024 kB
Writeback:           0 kB
AnonPages:       54624 kB
Mapped:          13356 kB
Slab:            13768 kB
SReclaimable:     8512 kB
SUnreclaim:       5256 kB
PageTables:       1288 kB
NFS_Unstable:        0 kB
Bounce:              0 kB
CommitLimit:   1569648 kB
Committed_AS:   393220 kB
VmallocTotal:   110584 kB
VmallocUsed:      9712 kB
VmallocChunk:   100824 kB
HugePages_Total:     0
HugePages_Free:      0
HugePages_Rsvd:      0
HugePages_Surp:      0
Hugepagesize:     2048 kB

afs-test:~ # cat /etc/openafs/cacheinfo
/afs:/var/cache/openafs:58000

afs-test:~ # rcopenafs-server start
Starting OpenAFS Server                                              done

Starting AFS cache scan...<4>printk: 3 messages suppressed.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
allocation failed: out of vmalloc space - use vmalloc=<size> to increase
size.
afs_osi_Alloc: Can't vmalloc 65536 bytes.
afsd:  memCache allocation failure at 95168 KB.
afsd: memory cache too large for available memory.
afsd: AFS files cannot be accessed.

[snip]

What's going on?





> it was hotly debated before, should be in the archives. sleep behavior, perhaps.
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info
>
> !DSPAM:1,4c374893257229074830735!
>
>
>
>   

-- 
--------------------------------------------------------------------------------
Claudio Prono                         OPST
System Developer               
                                      Gsm: +39-349-54.33.258
@PSS Srl                              Tel: +39-011-32.72.100
Via San Bernardino, 17                Fax: +39-011-32.46.497
10141 Torino - ITALY                  http://atpss.net/disclaimer
--------------------------------------------------------------------------------
PGP Key - http://keys.atpss.net/c_prono.asc





--------------050101090601000706020903
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Derrick Brashear ha scritto:
<blockquote
 cite="mid:AANLkTimV4guhmR_7ngnLX374mM40IkyPW7Z7XMbXqbjL@mail.gmail.com"
 type="cite">
  <pre wrap="">On Fri, Jul 9, 2010 at 12:03 PM, Andrew Deason <a class="moz-txt-link-rfc2396E" href="mailto:adeason@sinenomine.net">&lt;adeason@sinenomine.net&gt;</a> wrote:
  </pre>
  <blockquote type="cite">
    <pre wrap="">On Fri, 9 Jul 2010 11:51:01 -0400
Derrick Brashear <a class="moz-txt-link-rfc2396E" href="mailto:shadow@gmail.com">&lt;shadow@gmail.com&gt;</a> wrote:

    </pre>
    <blockquote type="cite">
      <blockquote type="cite">
        <pre wrap="">Could we not just try to kmalloc if vmalloc fails repeatedly?
        </pre>
      </blockquote>
      <pre wrap="">we don't elsewhere allocate sizes &gt; PAGESIZE from kmalloc, and this
definitely would be.
      </pre>
    </blockquote>
    <pre wrap="">Yes, but why not? kmalloc can alloc blocks bigger than PAGESIZE, I
thought (it goes up to 128k or something like that?).
    </pre>
  </blockquote>
  <pre wrap=""><!---->
  </pre>
</blockquote>
But... there is something strange...<br>
<br>
afs-test:~ # cat /proc/meminfo<br>
MemTotal:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1034864 kB<br>
MemFree:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 821272 kB<br>
Buffers:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 52100 kB<br>
Cached:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 79408 kB<br>
SwapCached:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0 kB<br>
Active:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 102700 kB<br>
Inactive:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 84524 kB<br>
HighTotal:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 131008 kB<br>
HighFree:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 252 kB<br>
LowTotal:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 903856 kB<br>
LowFree:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 821020 kB<br>
SwapTotal:&nbsp;&nbsp;&nbsp;&nbsp; 1052216 kB<br>
SwapFree:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1052216 kB<br>
Dirty:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1024 kB<br>
Writeback:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0 kB<br>
AnonPages:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 54624 kB<br>
Mapped:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 13356 kB<br>
Slab:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 13768 kB<br>
SReclaimable:&nbsp;&nbsp;&nbsp;&nbsp; 8512 kB<br>
SUnreclaim:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 5256 kB<br>
PageTables:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1288 kB<br>
NFS_Unstable:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0 kB<br>
Bounce:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0 kB<br>
CommitLimit:&nbsp;&nbsp; 1569648 kB<br>
Committed_AS:&nbsp;&nbsp; 393220 kB<br>
VmallocTotal:&nbsp;&nbsp; 110584 kB<br>
VmallocUsed:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 9712 kB<br>
VmallocChunk:&nbsp;&nbsp; 100824 kB<br>
HugePages_Total:&nbsp;&nbsp;&nbsp;&nbsp; 0<br>
HugePages_Free:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0<br>
HugePages_Rsvd:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0<br>
HugePages_Surp:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0<br>
Hugepagesize:&nbsp;&nbsp;&nbsp;&nbsp; 2048 kB<br>
<br>
afs-test:~ # cat /etc/openafs/cacheinfo<br>
/afs:/var/cache/openafs:58000<br>
<br>
afs-test:~ # rcopenafs-server start<br>
Starting OpenAFS Server&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
done<br>
<br>
Starting AFS cache scan...&lt;4&gt;printk: 3 messages suppressed.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
allocation failed: out of vmalloc space - use vmalloc=&lt;size&gt; to
increase size.<br>
afs_osi_Alloc: Can't vmalloc 65536 bytes.<br>
afsd:&nbsp; memCache allocation failure at 95168 KB.<br>
afsd: memory cache too large for available memory.<br>
afsd: AFS files cannot be accessed.<br>
<br>
[snip]<br>
<br>
What's going on?<br>
<br>
<br>
<br>
<br>
<br>
<blockquote
 cite="mid:AANLkTimV4guhmR_7ngnLX374mM40IkyPW7Z7XMbXqbjL@mail.gmail.com"
 type="cite">
  <pre wrap="">it was hotly debated before, should be in the archives. sleep behavior, perhaps.
_______________________________________________
OpenAFS-info mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OpenAFS-info@openafs.org">OpenAFS-info@openafs.org</a>
<a class="moz-txt-link-freetext" href="https://lists.openafs.org/mailman/listinfo/openafs-info">https://lists.openafs.org/mailman/listinfo/openafs-info</a>

!DSPAM:1,4c374893257229074830735!



  </pre>
</blockquote>
<br>
<pre class="moz-signature" cols="72">-- 
--------------------------------------------------------------------------------
Claudio Prono                         OPST
System Developer               
                                      Gsm: +39-349-54.33.258
@PSS Srl                              Tel: +39-011-32.72.100
Via San Bernardino, 17                Fax: +39-011-32.46.497
10141 Torino - ITALY                  <a class="moz-txt-link-freetext" href="http://atpss.net/disclaimer">http://atpss.net/disclaimer</a>
--------------------------------------------------------------------------------
PGP Key - <a class="moz-txt-link-freetext" href="http://keys.atpss.net/c_prono.asc">http://keys.atpss.net/c_prono.asc</a>



</pre>
</body>
</html>

--------------050101090601000706020903--